What problem does it solve? Debugging TypeScript backend incidents often devolves into guessing fixes, shotgun changes, or premature root-cause claims. This Skill enforces a disciplined root-cause investigation method that turns vague symptoms into a named failure surface, competing hypotheses, and the single best next diagnostic step. ## Core Features & Use Cases - Failure Surface Classification: Localizes incidents across Fastify lifecycle, Prisma/PostgreSQL, Redis, external HTTP integrations, streaming/workers, reliability, performance, and observability seams. - Hypothesis Discrimination: Maintains 2-4 competing mechanisms with explicit falsification criteria and picks the next probe with the highest discriminating power and lowest blast radius. - Root-Cause Quality Bar: Distinguishes triage plans, leading hypotheses, measurement gaps, and confirmed root causes before any fix direction is suggested. - Use Case: When a production API intermittently returns 503s under load, use this Skill to separate Prisma pool wait from slow SQL, retry amplification, and readiness misconfiguration before changing any timeout or pool setting. ## Quick Start Ask the AI to systematically debug why your Fastify endpoint intermittently times out, naming the failure surface, leading hypotheses, and the single next diagnostic step.
